1. Field of the Invention
The present invention relates to direct phase and frequency demodulation.
2. State of the Art
Much attention has been focused in recent years on the realization of radio systems using digital logic. Digital logic is less subject to manufacturing variability and is often much simpler and consumes lower power than an equivalent analog structure. In particular, current communications techniques, at least in wireless communications, are based in large part on quadrature modulation and demodulation. Such modulators and demodulators are relatively power-hungry and subject to various inaccuracies and limitations that become apparent as performance is pushed to higher levels.
One proposal for a generally-applicable angle demodulator is described in the dissertation by the present inventor entitled Extended Phase Shift Keying, deposited in the library of the University of California at Davis in August, 1998. Of particular interest is the so-called Time-Shift Angle Demodulator (TSAD), described in detail in Section 3.6.2 and Appendix B of the dissertation.
A block diagram of a TSAD 100 is shown in FIG. 1. The TSAD includes two stages, a time difference detector 101 and a pulse combiner/filter 103. The time difference detector measures time between adjacent the rising zero crossings of the two input signals s(t) and r(t). The first of these input signals is the angle modulated signal s(t). The second input signal is an applied reference signal r(t), which is assumed to be matched to the carrier of the angle modulated signal.
The timing information measured by the tune difference detector is provided on one of two outputs, a(t) and b(t), at any one time, as a fixed amplitude pulse with width equal to the timing difference of the input signals. The output selected depends on which input signal a rising zero crossing is observed first. If a rising zero crossing of the modulated signal s(t) occurs first, then the measurement output appears on the phase read output, a(t). If a rising zero crossing of the reference signal r(t) occurs first, then the measurement output appears on the phase lag output, b(t).
The time difference detector may be realized as a sequential phase-frequency detector (S-PFD), as shown in FIG. 2. The pulse combiner/filter may take the form of a differencing circuit followed by a low-pass filter (LPF).
Referring to FIG. 3, an example of operation of the TSAD is shown. The input signal s(t) is assumed to have a frequency 20% lower than that of the reference signal r(t). Initially, the phase of s(t) leads the phase of r(t). The two outputs from the time difference detector, a(t) and b(t), are shown. As the phase of s(t) shifts from leading to lagging the phase of r(t), a transition occurs in which pulses, instead of appearing in the output signal a(t), appear instead in the output signal b(t). The signals a(t) and b(t) are combined to form a difference signal a(t)-b(t), which is low-pass filtered, yielding an output signal v(t) that reflects the linear phase ramp relationship between the two input signals. Because the S-PFD aliases when phase shifts exceed 2π, the S-PFD output “jumps” following the accrual of 2π radian phase shift on the reference signal with respect-to the input signal. Corresponding jumps are reflected in the output signal v(t).
The transfer characteristic of the TSAD of FIG. 1 is shown in FIG. 4, showing the output v(t) as a function of the time difference d(t), where Te. represents the period of the reference signal. The transfer characteristic consists of overlapping line segments each having an extent of 2Tc, which corresponds to 4π radians of phase for the reference signal. When the time difference between the input signals becomes greater than 2π or less than −2π, a jump occurs to the next line segment as shown by the arrows in FIG. 4.
In order to increase the usefulness of the TSAD, it is necessary to eliminate the foregoing jumping characteristic, by shifting the line segments of FIG. 4 to form a single continuous line, as illustrated in FIG. 5. Such modification of the transfer relation may be accomplished using a “phase unwrapper,” a diagram of which is shown in FIG. 6. The phase unwrapper 610 includes a jump detection unit 611, a counter 613, a multiplier 615 and an adder 617. In operation, when a jump is detected, the appropriate signal offset, expressed as 2πK, is added or subtracted to form a corrected output signal vu.
Referring to FIG. 7, correction may also be achieved using phase scaling. Both the input signals are passed through respective 1/N frequency dividers 701, 703 before being applied to the TSAD of FIG. 1. In order for the input signal to effect a 2π phase shift at the input of the TSAD, there must be a 2πN phase shift at the input to the frequency divider. Thus, with respect to the original signal, the TSAD appears to have had its range extended by a factor of N.
Further simplification of the TSAD, while preserving the transfer relationship of FIG. 5, is desired.